DECISION MAKING
In its most simplistic sense, a decision is a choice among alternatives available to an
individual. It is the result of some consideration of facts and judgments that leads to a
specific course of action. The individual considers what is known and what is suspected to
select the alternative action that is most likely to bring a good outcome to that individual
or organization. As with most things, there is a range of difficulty of decisions from quite
simple and well structured at one end of the spectrum to what some refer to as wicked
problems at the other end. The tools to address the "simple" decision and alternatives that
should be considered are well understood and probably are similar to many other choices
that have been considered in the past. At the other end, the decisions are unique and quite
hard to formulate and often have no single correct answer and may not event have a good
answer. Generally DSS are not used to support the well-structured, easy problems. Rather,
they tend to be used for poorly structured, poorly understood problems for which neither
the solution nor the approaches to solving the problem are well understood.
Simon (1977) identified decision making as a three-step process as shown in Figure
2.1.
In the first step, intelligence, the decision maker
is
identifying
a
problem or opportunity.
To accomplish that task, the decision maker gathers information from the environment and
assesses the organization's performance in terms of the goals. This might be examining
how a particular organization is performing relative to others or examination of activities
within the organization and how they perform relative to expectations. It is at this stage
that business intelligence is particularly helpful to the decision maker. The second step
is design. In this step, the decision maker frames the particular choice to be made. He
or she establishes the specific objectives to be considered in a particular choice context
Decision Support Systems for Business Intelligence
by Vicki L. Sauter
Copyright © 2010 John Wiley & Sons, Inc.
24
DECISION MAKING
Figure 2.1. Nature of decision making.
and identifies appropriate alternatives. This step generally includes framing of alternatives,
collection of data, modeling, and examination of factors that might not fit into the model.
In the third step, the decision maker considers the information, compares alternatives,
selects the best alternatives, and evaluates that choice for its sensitivity to assumptions. The
goal of the DSS is to bring together appropriate business intelligence and models to help
that individual to consider a problem or opportunity from more perspectives with better
information.
To help the decision maker, the DSS needs to provide support in a number of areas.
First, the DSS must help decision makers identify and define the problem or opportunity.
Of course, this includes helping them see that a problem or opportunity exists, but it also
means helping them frame the problem or opportunity in terms of organizational objectives
and constraints and identify the appropriate people to be involved in the choice process.
Such framing of a choice helps decision makers to focus on the remainder of the steps
of the choice process. Second, DSS help decision makers identify alternative actions that
would address the problem or seize the opportunity. This requires the DSS to help identify
actions and to facilitate creative brainstorming to identify other alternatives. Third, the
DSS must help to collect appropriate information and access appropriate models to process
that information. The DSS must help decision makers process data, analyze data, and
determine how the data are actionable. Once alternatives are evaluated, the DSS must help
them examine their solution for its sensitivity to assumptions and the reasonableness of the
assumptions. Finally, after the decision is made, it is critical that the DSS help decision
makers monitor the results of the choice and assess the decision in terms of
the
process and
outcome. Said differently, the goal of the DSS is to help the decision maker make choices
better and more easily.
Such a goal is needed today more than ever. Decision makers have not only more
choices but also more complex choices every day. Some have access to automated tools,
but not all have what they need for each kind of decision. Further, a survey by Teradata
RATIONAL DECISIONS
25
reported that 70% of executives believed that "poor decision making is a serious problem
for business (Taylor and Raden, 2007).
Before we can discuss how to
support
the choice process, it is necessary to review what
we know about the choice process. The considerable amount of known information cannot
be chronicled here. Instead, we will take an overview of the general ideas about decision
making as they apply to
the
provision
of business intelligence and the
design
of
a
DSS.
The
guiding principle of
this
literature is that different decision makers will need quite different
information to support their choice processes. Similarly, a given decision maker will need
different support when facing different choices in different choice environments. Designers
of good DSS will be cognizant of those needs and respond to them so as to provide decision
makers with the flexibility to change the emphasis they place on various criteria.
RATIONAL DECISIONS
The place to begin is with a definition of
rationality.
Everyone knows that rational decisions
are better than those that are not rational. But what does "rational" mean? The dictionary
defines it as "based on, or derived from reasoning ... implies the ability to reason logi-
cally" (Guralink, 1980, p. 1179). Clearly, rational decisions require information about the
alternatives, which must be identified and evaluated with regard to some set of criteria and
some forecast of future conditions. In addition, we must judge these alternatives in terms
of their relative use of raw materials, their impact upon our constraints, and their benefits
in terms of our objective.
While this provides some guidance, it leaves a significant amount of room for inter-
pretation about what should be in a DSS. Rational decisions certainly are based
partly
on economic bases and therefore optimize the economic condition of the firm, such as
minimizing costs, maximizing profits, or maximizing return for investors. So, DSS need
In his book
The Pursuit of WOW/,
Tom Peters (1994, p. 74) discusses principles of management.
In principle 49, he notes how people respond to uncertainty:
The Greeks knew little of the way their world worked by the standards of Copernicus or
Newton, let alone Einstein. Yet they develope
d a system of meaning as finely articulated
as any you'll find in a modem quantum mechanics text.
The translation to everyday life is clear When confronted with anything unusual, from
a new ache or pain to a new boss, we try to build a theory of how things are going to
work out. And, says experience and psychological research, the less we know for sure,
the more complex the webs of meaning (mythology) we spin.
While Peters goes on to explain the lesson of keeping customers informed, this principle can have
other lessons to DSS needs. That is, without current and appropriate information and decision
aids,
decision makers will still develop a model of the choice context and make decisions based
on that model. With reasonable support and information
, decision makers arc likely to develop
a prudent model, Without reasonable support and information, decision makers are likely to
develop defective views of reality which can lead to imprudent choices being made. Hence,
decision support—even fairly limited support—can increase the likelihood of discerning choices
being made.
26
DECISION MAKING
Figure 2.2. Forms of rationality.
to be able to reflect how much each alternative will cost or how much profit will result
from each alternative. Consider, for example, the situation where a decision maker selects
a vehicle from a range of automobiles. Economic rationality would dictate that the costs
of the various automobiles be listed. In addition, also included might be more extensive
information such as the fuel mileage (so we could estimate the fuel costs during ownership),
the maintenance record (so as to be able to estimate maintenance costs), special insurance
issues (such as high theft rates or other attributes that raise the cost of insurance), and the
life expectancy of the automobile (so we would know when to replace the automobile).
Few of us can imagine purchasing or leasing an automobile without considering the price
in some way.
The clear importance of economic considerations means that DSS need to include
some economic data and models for evaluating those data. Unfortunately, since many
individuals overemphasize this criterion, many DSS are built to include
only
the economic
characteristics of the problem. However, just as few of us would consider buying a car
without some fiscal evaluation, few of us would consider
only
economic issues in the choice
process. In fact, as Figure 2.2 summarizes, there are six forms of rationality associated with
a reasonable decision process.
Upon reflection, almost everyone would agree that technical rationality is assumed
in a reasonable decision process. Technical rationality asserts that if the options will not
work, they should not be considered in the choice process. That is, choices should be
consistent with the attainment of our goals or objectives. For example, will a particular mix
of materials provide the needed strength or will a particular software package allow the
user to perform necessary computations? Even before we look at the economic benefit of
the system, we should ensure that the solution will actually solve the problem and meets
the needs of decision makers. Therefore, a DSS must include appropriate data and models
with which to evaluate the technical aspects of
the
choices. These might be the engineering
specifications of an alternative or information regarding the strength of materials relative
to needs. In addition, the system might incorporate a model for testing a design. Finally, it
might include a plan of action to meet some specific need, with references and information
about the success of such a plan in meeting needs in other locations.
RATIONAL DECISIONS
To return to our automobile example: What technical characteristics allow the decision
maker to decide whether or not the automobile would meet the needs of the owner? For
example, if the goal of the owner is high performance, technical criteria should include the
engine size, the horsepower, and the availability of possible options for improvement of
the performance, such as better grade wheels and tires. If instead the goal of the owner is
to be able to carry certain cargo or a certain number of passengers, then technical criteria
should include the type of trunk, the capacity of the trunk, the number of seats, and the
size of the automobile. Consumer report data, highway testing data, insurance data, and
other performance information might be relevant. The question of technical rationality is
whether the particular automobile will meet the specific needs of the user.
In most corporations, legal rationality, the third form of rationality in Figure 2.2, is
assumed in a reasonable decision process. Legal rationality prescribes that before a choice
is accepted, the decision maker(s) should ensure that the action is within the bounds of
legality in
the
jurisdiction in which the activity will take place. That
is,
if
the
manufacturing
process is to be completed in Indonesia, then the decision makers understand that the
process complies with the legal statutes of Indonesia as well as with those statutes of the
corporate headquarters and/or the country to which parts will be shipped. At the very least,
rationality would suggest that the decision makers be aware of the risk and implications of
violating statutes.
While most corporations evaluate the legal ramifications of a decision, few look at
the legal issues as an active component of the choice process. While decision makers
might share decisions with lawyers and ask their opinions, it is generally
after
most of
the generation of alternatives, trade-offs, and evaluation has occurred. Rarely is the legal
counsel enough a part of the decision-making team to participate actively in what-if kinds
of analyses. A DSS that will truly
support
the decision makers will provide access to data
and models through which to check the legality of the choices under consideration.
Consider again the choice of an automobile. The owner needs to guarantee that the
automobile of choice meets the legal requirements of the
state.
This might not
be
as straight-
forward as it appears at first glance. For example, suppose the owner wants to purchase a
preowned automobile, and suppose the system's database includes many automobiles man-
ufactured before 1970 when seat belts were not required on U.S. automobiles, including
many "classic" and antique cars. The law does not prescribe that these cars be retrofitted
with seat belts, so there is no legal issue associated with the purchase of the car. However,
there may be a legal issue associated with the use of
the
car if, for example, the owners have
small children who will ride in it. Car seats, which are required in many states, cannot be
secured properly without seat
belts.
Hence, if
the
owners purchased a "classic car" (or even
an antique car) that had not been retrofitted with seat
belts,
the children could not ride in the
car legally because their car seats could not be secured in the back seat. If the owners were
not familiar with the seat belt law, they might not consider this issue until after they had
already purchased the car. However, if the DSS truly provided support, it would provide
users such information about legal issues as they were narrowing down alternatives. Most
decisions have some legal issues that should be considered during the decision process.
Social rationality is a consideration of the ethical nature of the choice from the per-
spective of both the society as a whole and the decision unit as a group. It suggests that
decision makers will not make choices that are "good for the company" if they are bad for
themselves or their department. Similarly, decision makers will not select an option if it is
in conflict with the prevailing mores of society. Consumers increasingly expect companies
to be socially responsible in their actions, and companies are responding with corporate
plans and annual social responsibility reports. Where such plans and reports are available,
they should be integrated into a DSS to help decision makers assess social responsibility.
